Built for the data your other databases can't touch.
Hybrid search. Local ML inference. Multimodal documents. One binary, zero glue code. Free to run in swarm mode, ready to scale with Antfly Cloud.
One database for documents, vectors, and knowledge graphs.
Stop stitching together a vector database, a document store, an embedding service, and a reranker. Antfly handles it all — distributed by design, with built-in ML inference powered by Termite.
Hybrid Search
BM25 keyword search and vector similarity in a single query. No external services, no glue code.
Multimodal Documents
Index PDFs, images, audio, and video alongside text. Antfly extracts, chunks, and embeds automatically.
Local ML Inference
Built-in Termite engine for embedding, reranking, and chunking. No external API calls required.
Distributed by Design
Raft consensus, horizontal scaling, and automatic rebalancing. Start with one node, scale to many.
Knowledge Graphs
First-class graph relationships between documents. Traverse connections, not just similarity.
Developer Friendly
REST API, TypeScript SDK, Python SDK. Works with LangChain, LlamaIndex, and CrewAI out of the box.
Get Antfly running in seconds:
# Install and start in swarm mode (single-node, free)
curl -sSL https://antfly.io/install | sh
antfly start --mode swarm
# Create a collection and index a document
curl -X POST http://localhost:7700/collections \
-H "Content-Type: application/json" \
-d '{"name": "docs", "embedding_model": "default"}'
curl -X POST http://localhost:7700/collections/docs/documents \
-H "Content-Type: application/json" \
-d '{"content": "Antfly is an AI-native database."}'
# Search with hybrid (BM25 + vector) ranking
curl "http://localhost:7700/collections/docs/search?q=what+is+antfly"Antfly Cloud
Multi-node, managed, and monitored. The same Antfly you run locally, deployed and scaled for you.